Marijuana and alcohol increase crash avoidance reaction time in a driving simulator test at blood concentrations below commonly-used per se ‘Cut-offs’ for Intoxication

Author(s): Kawthar Alali, Jessica Stewart, Rimzim Taneja, Edison Nwobi, Brianna Murdock, Tyiesha Head, Ki-Jana Malone, Mohammed B Mohammed, MaryAnne Stewart, Jamie McQueen, Doreen Head and Randall Commissaris* The present study demonstrates marijuanaand alcohol-induced impairment of a driving-relevant measure in a driving simulator task at (estimated) blood alcohol and THC concentrations that are below the per se cut-off for impaired driving in several states. The subject was an adult male with a history of occasional alcohol use (2-3 times/week for the past 6 months) and ... Abstract View Full Article View

recently moved to a 50mg/dl cut-off for drunk driving [2]; this value is in agreement with many other countries in the world, where the drunk driving cut-off also has been lowered to 50 mg/dl within the past 20 years [3]. The move to reduce the BAC cut-off for drunk driving is in part the result of research demonstrating a signifi cant crash risk with BAC levels 50 mg/ dl. NHTSA has advocated for reducing the BAC cut-off from 80 mg/dl to 50mg/dl across the US [4].
Marijuana continues to be the most commonly used illicit drug in the USA, with over 20 million people in the US reporting that they used marijuana within the past month [5]. With medical and legal marijuana continuing to be approved in many states, the frequency of use is likely to increase further [6]. Bonar recently reported that more than 50% of medical marijuana patients in the state of Michigan admitted to driving within one hour of use, and nearly 20% reported driving while high [7]. It has been reported that legalization of cannabis is associated with increased reports of cannabisrelated crash fatalities [8], although other studies have reported no signifi cant change as a consequence of legalization [9]. Similarly, reports on the effects of marijuana on driving in general are somewhat inconsistent [10]. Conducted a metaanalysis and reported that the risk of being involved in a crash signifi cantly increased after marijuana use. Marijuana also has been shown to impair driving simulator performance [11,12]. In contrast, a large case-control study conducted by the National Highway Traffi c Safety Administration found no signifi cant increased crash risk attributable to cannabis after controlling for drivers' age, gender, race, and presence of alcohol [13].
Compared to alcohol, there is less consensus regarding the establishment of a legal cut-off for marijuana-impaired driving. In states where marijuana is not legal for medical or recreational use, any amount of marijuana in the blood is deemed illegal.. In those states where marijuana is legal for medical or recreational use, different states use different cutoffs: 1, 2, or 5 ng/ml; some states with recreational and/or medical marijuana have no 'cut-off' concentration [14].
The wide variability between states regarding the identifi cation of a blood THC concentration for driving impairment, and the fact that some states have opted NOT to have a THC cut-off, relates to some of the challenges/problems of identifying the relationship between blood concentrations and effects of THC. This problem relates at least in part to the complicated pharmacokinetics of THC in the blood. THC exhibits biphasic pharmacokinetics. After reaching a peak (125-150 ng/ml) within minutes of inhalation, THC concentrations decline rapidly to approximately 10-20 ng/ml within 60-90 minutes (mostly via redistribution), followed by a much slower elimination phase, with a half-life of approximately 24 hours [15][16][17]. In addition, THC is sequestered in body fat and there is the potential for redistribution from fat back to the bloodstream [18]. The pharmacokinetics of orally administered THC are signifi cantly different from smoked marijuana, with a much lower peak concentration (5-10 ng/ml for 20 mg oral) and a later onset to peak concentration of 1-2 hours [16,17,19].
Another factor which complicates the issue of establishing a cut-off for THC concentrations and driving impairment is the potential contribution of 11-OH THC, an active metabolite of THC [20]. All of these issued complicate signifi cantly the question of establishing a legal cut-off for THC-impaired driving.
Moderate to heavy use of regular marijuana is likely to be associated with signifi cant tolerance [21,22]. Because We describe herein a driving simulator task that measures a defensive driving behavior in response to a surprise event, i.e., swerving to avoid an imminent crash. We further report the results of an open label study on the effects of marijuana on driving simulator performance in a subject with a known history of very limited marijuana use. For reference purposes, the study also included driving simulator assessments following a challenge with alcohol. We believe this is the fi rst-ever study of the effects of marijuana on driving simulator performance in a subject with a life history of virtually no marijuana use (< 10 lifetime uses) and absolutely no history of driving following marijuana use. We report the effects of alcohol and marijuana on this crash avoidance reaction time response, and we provide data comparing blood alcohol concentrations (BACs), THC and metabolite concentrations with driving simulator performance.

Subject
An adult male subject with a history of very limited marijuana use (fewer than 10 lifetime exposures; none in the 8 weeks prior to testing) and occasional alcohol use (less than 3 occasions/week and less than 3 drinks/occasion for the past 12 months); the subject has never had a DUI or DUID; the subject was not taking any CNS active drugs (legal or illegal) at the time of the study; the subject was experienced (more than 10 driving experiences) with the driving simulator in general and the crash avoidance procedure (more than 4 driving experiences).

This study was approved by the Wayne State University Internal
Review Board (WSU IRB #066716B3E).

Apparatus and crash avoidance testing procedure
The studies were conducted using a fi xed-base driving Citation: Alali  The subject was instructed to keep his eyes on the roadway and not on the speedometer, with driving speed advice/coaching (slow down a little; speed up a bit) provided by a laboratory team member so the driver would not check the speedometer to maintain driving speed. For each test session, videotaping of the 'roadway' and the driver's face were used to create synchronized picture-in-picture videos, which were viewed by an observer who was unaware of the treatment condition and scored each trial by assessing whether the driver's eyes were or were not on the roadway at the moment when the 'stalled car' fi rst appeared.  [24]. Estimates of whole blood THC, 11-OH-THC and THC-COOH were calculated by multiplying the respective plasma concentration by a factor of 1.6 [25].

Beer goggles control experiment
In a separate experiment, the subject performed two separate crash avoidance tests in a 'no drug' condition, once looking at the roadway at the moment the stalled car appeared. Data from crash avoidance trials were excluded from statistical analysis if this review revealed that the driver was not looking at the roadway (e.g., speedometer check) at the moment when the 'stalled car' appeared. Overall, approximately 5% of the trials were excluded for this reason. The criterion for statistical signifi cance was p<0.05 in all analyses.

Marijuana effects
Marijuana 'Pre-treatment' driving data was unobtainable due to a computer error. However, both 'Pre-treatment' and '24 hour Post-treatment' values measured in the alcohol study (see below) were comparable to '24 hr Post-treatment' of the marijuana study; therefore, the '24 hr Post-treatment' measure was used as the primary point of reference for statistical comparison. Figure 1 illustrates that marijuana signifi cantly increased crash avoidance reaction times (F[6,105]=6.93, p<0.001). Post hoc Student Neuman Keuls (SNK) analysis revealed signifi cant increases in reaction times at 120, 150 and 180 minutes when compared to the 24-hour 'post-treatment' test.
Not including the 24-hour post treatment THC blood value of zero, estimated THC blood concentrations ranged from 1.5 ng/ml to 2.9 ng/ml throughout the study. 11-OH THC estimated blood concentrations were slightly higher than parent THC at the various time points, and estimated blood THC-COOH concentrations were 5-10 times higher than THC, and it was present in measureable quantities before either THC or 11-OH-THC. The signifi cant increase in crash avoidance reaction time was observed during the time interval 120-180 minutes post ingestion when THC blood concentrations were estimated to be 2.9 ng/ml and 1.5 ng/ml respectively, and 11-OH-THC concentrations were estimated to be 3.1 ng/ml and 2.1 ng/ml, respectively, with subject reports of feeling 'high' (10/10). In contrast to the alcohol challenge (see below), there was no incidence of yawning observed throughout any test sessions for the marijuana experiment.

Discussion
In an experienced driver with a history of occasional alcohol use and very infrequent marijuana use, consumption of each drug signifi cantly increased crash avoidance reaction times.
Conversely, wearing 'beer goggles' dramatically affected visual performance tasks (walk a straight line; catch a ball), but did not affect crash avoidance reaction time. These data suggest that the crash-avoidance reaction time procedure described above using a fi xed-base driving simulator is a reliable and sensitive tool for studying the effects of alcohol, marijuana and perhaps other drugs (prescription, non-prescription and illegal) on reaction time in a defensive driving performance task. Future work with laptop, gaming chair or even virtual reality applications might allow for more widespread applications of this crash avoidance task for the study of drugs on this important defensive driving maneuver.
Signifi cant impairment of crash avoidance behavior was observed during the driving test at 60 minutes post-alcohol, where the BAC (estimated from plasma) ranged from 39 mg/ dl shortly before testing to 67 mg/dl shortly afterward. This fi nding is consistent with the argument for decreasing the legal BAC in the US from 80 mg/dl to 50 mg/dl [4], a move which has already been undertaken by many countries [14].
There is less clarity regarding a cut-off value for marijuana impaired driving for many reasons. In the present study, signifi cant impairment as exhibited by increased crash avoidance reaction time was observed when the whole blood THC concentrations (estimated from plasma) were 2.9 ng/ ml (125 minutes) and 1.5 ng/ml (190 minutes). These values are signifi cantly lower than the effect-based threshold values of 13.1 ng/ml (to mimic 80 mg/dl EtOH) and 8.2 ng/ml (to mimic 50 mg/dl EtOH) as reported by Hartmann, et al. [11] in a driving simulator study. There are several possible reasons for this difference. First, the primary measure in the Hartman, et al. [11] study was maintaining lane control as measured by standard deviation of lane position (SDLP), which might be less sensitive to the effects of marijuana when compared to crash avoidance. Second, it is possible that THC tolerance is responsible for the difference. The subject in the present study was a very infrequent user, virtually a novice, whereas in the Hartman, et al. [11] study the subject histories ranged from occasional use (<1x/mo) to relatively frequent use (2-3 x/wk). Chronic marijuana use produces signifi cant tolerance for a number of effects [21,22]. Consistent with this idea, preliminary results from studies in medical marijuana patients Figure 2: Plotted are the Mean + SEM (n=18-20 trials) Crash Avoidance Reaction Times (in msec) before (Pre) and at various times after consumption of alcohol. Alcohol was administered in two 'doses' of 3 beers each over a 15-minute period; one dose was administered immediately after the Pre-test, and the second dose was administered at approximately 50 minutes after the Pre-test. Vertical arrows indicate the times at which blood samples were collected. Ethanol concentrations were measured in plasma, and blood ethanol concentrations were estimated using the correction factor of 1. 16  (virtually all of whom are chronic marijuana users) suggest that they are indeed less affected by marijuana administration when compared to the subject in the present study (Alali et al., in preparation). Finally, the route of administration and associated pharmacokinetics in the two studies -oral in the present study and smoked in the Hartmann, et al. [11] studymight account for the differences observed in the effect-based threshold value [11,17,19]. All of the above point to the challenge of identifying a single cut-off value for marijuana-induced driving impairment.
Consistent with previous reports, plasma concentrations of the inactive THC metabolite (THC-COOH) were consistently higher than concentrations of the parent THC compound, and were measurable before THC parent concentrations were measureable [27]. It should also be noted that 11-OH-THC concentrations closely paralleled those of the parent THC; this is consistent with previous reports following oral THC administration [16,17,19]. Given the signifi cant biological activity of 11-OH-THC [20], these fi ndings suggest that measurements of this metabolite might also need to be considered when discussing a possible per se 'cut-off' for marijuana-intoxicated driving.
In summary, the results of the present studies suggest that the crash-avoidance reaction time procedure described above is a robust, reliable and sensitive tool for studying the effects of, and possible tolerance to, marijuana, alcohol and other drugs (prescription, non-prescription and illegal) on defensive driving performance.

Limitations of the present study
The present study used an open label design with a single subject; however, the present fi ndings are robust and statistically reliable. The crash avoidance reaction test procedure exhibits high test-retest reliability within subjects and is therefore suitable for multiple tests with different treatments, thus making longitudinal studies in a single subject highly valuable. The present study was conducted in a driving simulator and not on a real road under real driving conditions; however, driving simulator studies have been shown to be a very safe and effective tool for studying, understanding and predicting future real world driving experiences. Finally, there was not much 'open driving' in the study to examine for possible differential effects of the various drugs on driving behavior (e.g., lateral control, speed, aggression, risk-taking); however, the degree of simplicity and standardization used in the present study was important for the purpose of obtaining multiple crash avoidance reaction trials.

Declarations
Ethics approval and consent to participate: The study was conducted under Wayne State University (WSU) Internal Review Board (IRB) Approval #066716B3E; the subject provided written informed consent.